Download This PDF File

International Journal of Communication 14(2020), 5055–5071 1932–8036/20200005 Challenges in Codifying Events Within Large and Diverse Data Sets of Human Rights Documentation: Memory, Intent, and Bias JEFF DEUTCH1 Syrian Archive, Germany This article discusses challenges in codifying events within large and diverse data sets of human rights documentation, focusing on issues related to memory, intent, and bias. Clustering records by events allows for trends and patterns to be analyzed quickly and reliably, increasing the potential use of such content for research, advocacy, and accountability. Globally, archiving and preservation of user-generated digital materials documenting human rights abuses and war crimes are increasingly recognized as critical for advocacy, justice, and accountability. For the conflict in Syria, which began in 2011, there are more hours of user-generated content documenting rights violations uploaded to digital platforms than there have been hours in the conflict itself. Whereas some content has been clustered around specific events, such as larger open-source investigations by civil society and documentation efforts, the vast majority of content currently exists as individual unstructured records rather than jointly as clustered events within a relational database. The sheer amount of content and the near constant removals of materials from public channels mean that human rights monitors are in a race against time to preserve content, identify violations, and implicate potential perpetrators. Overcoming challenges related to memory and bias is crucial to this process. Keywords: preservation, open-source technology, open-source investigations, user- generated content, incident-based clustering, metadata, structured data, archiving, workflows, Syria Now entering its 10th year, the conflict in Syria has produced a massive amount of documentation concerning human rights violations and other crimes committed against the Syrian people, much of it Jeff Deutch: [email protected] Date submitted: 2017‒12‒04 1 The documentation preserved, verified, and analyzed by the Syrian Archive has been collected by groups and individuals living or working in Syria. This would not be possible without their work documenting and reporting war crimes and human rights violations. I am thankful to those who have assisted in developing and refining this article, in particular members of the Syrian Archive team including Libby McAvoy, Hadi al Khatib, and Abdul Rahman al Jaloud. Additional thanks to the International Journal of Communication and the anonymous reviewers for their insightful comments. Finally, I would like to thank Dima Saber for organizing the Image Activism After the Arab Uprisings symposium. Copyright © 2020 (Jeff Deutch). Licensed under the Creative Commons Attribution Non-commercial No Derivatives (by-nc-nd). Available at http://ijoc.org. 5056 Jeff Deutch International Journal of Communication 14(2020) uploaded to the Internet. Syrian human rights defenders have a preference for using social media platforms for publishing and publicizing documentation or at least use the medium effectively and often (Sterling, 2012). Large social media platforms such as YouTube, Facebook, and Twitter have become “accidental archives” never intended to host such content. Yet, this open-source content often has the potential to offer counternarratives to misinformation and propaganda published by actors on all sides of the conflict. When citizen journalists upload verifiable video footage of an event, that footage might disprove claims made by those in traditionally powerful positions, such as states. But unfortunately, for various reasons, much of this important documentation has been erased or underutilized in advocacy and accountability work (al Jaloud, al Khatib, Deutch, Kayalli, & York, 2019). To address this challenge, the Syrian Archive was founded in 2014 and has worked to preserve more than 3 million records of digital content from more than 3,000 sources, using this documentation to publish a number of investigations and long-form reports. The group aims to use this documentation to create a counternarrative to the misinformation created by all actors involved in the conflict in Syria, and to respond to human rights violations by introducing better tools, replicable methodologies, and a publicly available verified database for use in investigative journalism and reporting, evidence-based advocacy campaigns, and criminal case building to demand accountability against perpetrators of those violations. Losing this documentation may directly affect justice and accountability efforts by Syrian regional and international civil society organizations. In some cases, this documentation might offer the only evidence that a war crime has happened. In an interview with The Atlantic, Syrian Archive founder Hadi al Khatib spoke to the ways in which losing this documentation will make it harder for groups to do their work. He stated, [Cracking down on so-called extremist content deemed off limits in the West] could have ripple effects that make life even harder for those residing in repressive societies, or worse, in war zones. Any further crackdown on what people can share online . would definitely be a gift for all authoritarian regimes. On the ground in Syria . Assad is doing everything he can to make sure the physical evidence [of potential human rights violations] is destroyed, and the digital evidence, too. The combination of all this—the filters, the machine-learning algorithms, and new laws—will make it harder for us to document what’s happening in closed societies. (Warner, 2019, para. 19) Although preserving this content is important for justice and accountability efforts, the removal or erasure of this content also risks destroying the collective Syrian digital memory formed since 2011, which may result in ignoring violations committed by all parties in the conflict and prevent future generations from knowing what happened in their country. Existing research by Syrian Archive researchers describes the data models used for storing and processing large data sets of human rights documentation (Deutch & Para, 2020). This includes processes of transforming diverse types of content from “original” online formats to “archival” formats accessible to researchers. International Journal of Communication 14(2020) Challenges in Codifying Events 5057 Further research by Syrian Archive researchers describes the operational models used to verify individual records (Deutch & Habal, 2018) for use in investigative pieces. In the case of videos uploaded to YouTube, for example, this involves knowing the source of a video, where a video was filmed, and when a video was filmed. Additional information regarding a video, such as landmarks present, the time of day, munitions observed, and others, aids in this process. In this article, I discuss challenges related to archiving large and diverse data sets of human rights documentation, using the case study of the Syrian Archive’s clustering of individual records into a relational database of “events.” In the process, I focus on issues related to memory, intent, and bias. With millions of records preserved, contextualizing content into events would add an additional layer and increase the collection’s potential to support and further justice and accountability efforts of all kinds. Clustering records by events allows for trends and patterns to be analyzed quickly and reliably, increasing the potential use of such content for research, advocacy, and accountability. Furthermore, clustering content into events could aid in identifying potential perpetrators as well as intent, both useful for advocacy and accountability. I first discuss the use of imagery and “emotive affectivity” by advocacy organizations to pursue social justice aims, as well as data loss relating to records documenting human rights violations committed in the Syrian conflict. I then discuss the importance of archiving these records and the potential use of these records in legal proceedings. Finally, I discuss challenges in contextualizing these records, using the example of the Syrian Archive’s clustering records into event-specific databases, as well as issues relating to bias of which types of violations are documented. Memory and Machines There often exist tensions between authenticity and reliability of machine and human memory. Speaking to these tensions, Garcia (2016) writes, Both human memory and machine memory are subject to the limitations, vulnerabilities, and maintenance requirements of its material medium, whether it is soft flesh or a hard drive. Everything fails, sooner or later. And, as foreseen in even the earliest of cyberpunk literature, machine memory is not inherently more truthful . it may be more precise, dense, and enduring in its recall of stored information, but the quality of that data is uncertain from the outset and subject to tampering and corruption. (p. 90) In 1982, TIME Magazine’s “Man of the Year”2 was a machine: the computer. The title, which is given annually to “the greatest influence for good or evil,” was meant to mark the start of the information age (“The Computer Moves in,” 1983). That same year, Philip K. Dick’s (1968) most famous novel Do Androids Dream of Electric Sheep? was used as the basis for Ridley Scott’s science fiction noir-thriller Blade Runner (Deeley & Scott, 1982). The film, set in 2019 in a dystopian futuristic Los Angeles, follows the story of blade runner Rick Decker,

Load more